A New Feature Selection Method Based on Fuzzy Updated Particle Swarm Optimization

Feature selection is one of the important problems in classification that has an important role in increasing efficiency and there are different methods to solve it. Particle swarm optimization is one of the algorithms based on swarm intelligence that has been used in different contexts including feature selection and has shown good performance. Many studies have used particle swarm optimization for feature selection. In a research accomplished in the field, the authors have presented several different strategies for initialization of particles and several methods to update personal best and global best in particle swarm optimization for feature selection and have achieved good results. In this article we have presented a method for feature selection based on the mentioned research and our proposed fuzzy updating for one of the personal best or global best. k-nearest neighbor is used as the classifier. Experiments is performed on several datasets. According to the done simulations, the proposed method obtains good results in terms of accuracy and the number of feature in comparison with reference article.